Get Complete Project Material File(s) Now! »
Current understanding of the passivation and transport mechanisms
Overall, the different passivation mechanisms involved in the poly-Si contact are the following:
▪ The thin SiOx layer providing chemical passivation of the c-Si surface.
▪ The high doping density of the poly-Si layer providing field-effect passivation.
▪ The additional hydrogen diffusion likely providing chemical passivation of the residual defects at the SiOx/c-Si interface (and potentially in the c-Si).
The addition of these three mechanisms results in the aforementioned excellent surface passivation properties. However, the final passivation provided by the poly-Si contact has been shown to strongly depend on its fabrication process (e.g. the SiOx growth or the annealing step). In the following we will address the current understanding of the interplays between the different process steps and the passivation mechanisms involved.
Growth of the SiOx layer
The SiOx growth is the first important step to control the final passivation properties of the poly-Si contact. More particularly, the SiOx layer has been shown to play a role on the thermal stability of the surface passivation during the following annealing step. Already in the 80s, Wolstenholme et al. observed by TEM that submitting the poly-Si contact to an excessive thermal budget led to a significant break-up of the SiOx layer [75]. More recently, other groups observed by TEM the breaking-up of the SiOx after annealing at excessive Ta [24,76]. They correlated it with high interface recombination current density (J0) values resulting from the increasing area of direct contact between the poly-Si layer and the c-Si substrate where the continuity of the SiOx layer is disrupted.
Moldovan et al. also emphasized through XPS analyses the link between the SiOx stoichiometry and the SiOx growth technique used [45]. They demonstrated that thermally grown SiOx are O-richer and show better surface passivation properties than their chemically grown counterparts. Poly-Si contacts featuring a thermal SiOx layer show also more stable surface passivation properties upon annealing at high Ta [76–78].
Role of doping
A second important phenomenon for controlling the passivation properties is the diffusion of dopants from the poly-Si layer into the c-Si. It is well known that dopants diffuse from the poly-Si layer to the c-Si substrate during the annealing step, the length of the diffusion tail increasing with increasing Ta [19,21,79,80]. The diffusion of dopants has been shown to depend on the thickness and the method used to grow the SiOx layer [78,81,82]. In particular, thermal SiOx layers have been shown to better block the diffusion of dopants in the c-Si, which could result from their O-rich stoichiometry leading to a denser SiOx phase compared to their chemical counterparts. Moreover, a shallow diffusion of dopants of a few tens of nanometers in the c-Si has been shown to improve the surface passivation properties of the poly-Si contact, which could result from an enhanced field-effect passivation.
The doping type of the poly-Si contact also plays a role on the final passivation properties. The n+-type poly-Si contacts (doped with P) historically demonstrated better performances than their p+-doped counterparts (doped with B). This difference could result from a higher segregation coefficient of B in SiOx phases compared to P, leading to B atoms piling-up at the SiOx interface and increasing the density of interface states (Dit) [16]. Recently, Young et al. demonstrated excellent surface passivation with p-type poly-Si contacts doped with Ga, which tends to confirm that B is indeed responsible for the lower surface passivation of poly-Si(p+) contacts doped with B [16].
Optimization of the annealing temperature
The annealing step has been observed to involve the degradation of the thin SiOx layer and the diffusion of dopants in the c-Si substrate. Therefore, maximizing the surface passivation properties arises from a careful optimization of the annealing temperature (Ta) to preserve the chemical surface passivation provided by the SiOx layer while enabling a shallow diffusion of dopants in the c-Si to enhance the field-effect passivation.
Many groups investigated the effect of Ta on the passivation properties of the poly-Si contact [33,52,79,83,84]. They usually observe a similar trend of iVoc versus Ta: iVoc increases with increasing Ta up to an optimal annealing temperature Ta,opt. Then, iVoc decreases with further increasing Ta above Ta,opt. The underlying phenomena that gives rise to this iVoc versus Ta trend are not formally established yet but some hypotheses have been proposed:
▪ The increase of iVoc with increasing Ta up to Ta,opt could result from an enhanced field-effect passivation due to the shallow diffusion of dopants from the poly-Si layer to the c-Si.
▪ The iVoc drop for Ta > Ta,opt could result from the SiOx disruption at too high Ta.
As mentioned before, recent studies emphasized the link between the thermal stability of surface passivation and the oxidation process used to grow the SiOx layer, which tends to confirm that the SiOx layer is at least partly responsible for the iVoc decrease for Ta > Ta,opt. Excessive dopant diffusion in the c-Si is also sometimes proposed to explain the iVoc drop for Ta > Ta,opt, as it would enhance Auger recombination in the highly-doped c-Si.
Role of hydrogen
As mentioned, a hydrogenation step is generally performed after annealing of the poly-Si contact to enhance the final surface passivation properties. While the benefit of the hydrogenation step on passivation has been observed at a macroscopic scale (i.e. J0 decrease and iVoc increase), a fine understanding of the mechanisms involved is still lacking.
A common hypothesis to explain the passivation gain is that H diffuses up to the SiOx/c-Si interface and reduces interface defect density (Dit). Several studies reported SIMS measurements of the poly-Si contact after hydrogenation showing the accumulation of H (or D) at the SiOx interfacial layer [16,85,86]. Interestingly, Lehmann et al. did not observe this H accumulation without a SiOx layer intentionally grown at the poly-Si/c-Si interface. Moreover, they observed the H accumulation only at the interfaces of a 25-nm thick SiOx grown in between the poly-Si layer and the c-Si substrate.
Another interesting observation is that intrinsic poly-Si contacts benefit as much from the hydrogenation as their highly-doped counterparts [85]. Moreover, the passivation gain after hydrogenation was shown to remain if the H-rich dielectric layer or even the poly-Si layer are etched-off [21,86]. All together, these observations tend to prove that H chemical passivation on its own provides an excellent passivation of the SiOx/c-Si interface. However, the stability of the chemical passivation provided by H could be a matter of concern as the H is suspected to out-diffuse from the poly-Si contact in case of following high-temperature process steps performed especially after etching-off the H-rich dielectric layers [87,88].
Table of contents :
General introduction
Part 1. State-of-the-art
I. Towards higher passivation level for c-Si solar cells
1. Working principle of conventional c-Si solar cells
2. Main loss mechanisms of a c-Si solar cell
3. Concept of full-area passivating contacts
II. Poly-Si contact fabrication and integration in solar cells
1. Brief historical review
2. Working principle of the poly-Si contact
3. Fabrication process of the poly-Si contact
a. Growth of the thin SiOx layer
b. Si-based layer deposition and doping
c. Annealing step
d. Hydrogenation step
4. Integration of the poly-Si contact in c-Si solar cells
a. Record efficiencies on small area solar cells
b. Towards industrial solar cells integrating the poly-Si contact
i. The metallization challenge
ii. Large area c-Si solar cells integrating the poly-Si contact
III. Current understanding of the passivation and transport mechanisms
1. Control of the final surface passivation
a. Growth of the SiOx layer
b. Role of doping
c. Optimization of the annealing temperature
d. Role of hydrogen
2. Transport of charge carriers through the SiOx layer
Part 2. Experimental
I. Fabrication process of B-doped poly-Si/SiOx contacts by PECVD
1. Si-based layers by Plasma-Enhanced Chemical Vapor Deposition (PECVD)
2. c-Si substrates and samples configuration
3. Fabrication of in-situ doped poly-Si/SiOx contacts
4. SiOxNy:B layers for ex-situ doping of the poly-Si/SiOx contacts
II. Characterization techniques
1. Structural and chemical characterization
a. Thickness and microstructure of the layers
i. Spectroscopic ellipsometry
ii. Transmission Electron Microscopy
b. SiOx chemical composition: X-ray Photoelectron Spectroscopy
c. Doping profiles in a-Si:H and poly-Si layers
i. Electrochemical Capacitance-Voltage
ii. Secondary Ions Mass Spectroscopy
2. Electronic characterizations
a. Electrical properties of the poly-Si layer: Hall effect technique
b. Lifetime measurements: Photo Conductance Decay
c. Defect characterization in c-Si: lifetime spectroscopy
i. SRH description of an electrically active defect center
ii. Lifetime spectroscopy methods
iii. Case of several SRH defect centers
iv. Bulk and surface contributions to the effective lifetime
v. SRH formalism at the surface
d. Transport of charge carriers in the poly-Si/SiOx contact: Conductive Atomic Force Microscopy
Part 3: Fabrication of the poly-Si/SiOx contact
I. Optimization of the in-situ doped poly-Si/SiOx contact
1. Optimization of the deposition conditions for blister-free poly-Si(B) layers .
a. Effect of the deposition temperature increase
b. Effect of the silane dilution in dihydrogen
c. Evaluation of the poly-Si electrical properties through deposition optimizations
d. Effect of the blistering on the stability of surface passivation
2. Crystallization and doping of the poly-Si layer
a. Crystallization of the poly-Si layer
b. Doping of the poly-Si layer
3. Surface passivation properties of the in-situ doped poly-Si(B)/SiOx contact
a. Impact of the annealing temperature
b. Impact of the hydrogenation step
II. Ex-situ doping of the poly-Si/SiOx contact
1. Deposition of the intrinsic Si layer by PECVD
a. Decrease of the in-situ B-doping
b. Optimization of the intrinsic Si layer
2. Ex-situ doping
a. Si layer microstructure
b. Comparison of the two ex-situ doping processes
i. Electrical properties
ii. Passivation properties
c. Doping profile of the ex-situ doped poly-Si/SiOx contact
3. Comparison of in-situ and ex-situ doped poly-Si/SiOx contacts
a. Si-layer microstructure
i. After deposition
ii. After annealing
b. Electrical properties
c. Surface passivation properties
4. Effect of the doping density and hydrogen diffusion on surface passivation
a. Passivation stability upon annealing and SiN:H deposition
b. H diffusion profile in poly-Si(i) and poly-Si(B) contacts
c. Effect of the firing temperature
Part 4. Investigation of the poly-Si/SiOx interface
I. Chemical and structural evolution of the SiOx layer
1. SiOx grown at the c-Si surface
a. Ex-situ annealing in the tube furnace
b. In-situ annealing in the XPS chamber
c. Quantitative comparison of the XPS spectra after annealing
2. XPS study of the poly-Si/SiOx/c-Si stack
a. Qualitative study of the oxides upon annealing
b. Quantitative analysis
c. Structural evolution of the SiOx at the poly-Si/c-Si interface
3. Conclusions on XPS and TEM analysis of the interface
II. Investigation of the transport mechanism within the SiOx layer
1. Impact of the surface preparation of the sample on the C-AFM measurement
2. C-AFM investigation of the pinholes formation within the SiOx layer
a. Transversal measurements
i. With and without interfacial SiOx
ii. Comparison of chemical and thermal interfacial SiOx layers
b. Lateral C-AFM measurements
3. KPFM investigation of the pinholes formation within the SiOx layer
III. Study of the poly-Si/c-Si interface recombination by lifetime spectroscopy
1. Sample preparation
2. Extraction of the surface recombination velocity
3. Linearization of τsurf using Murphy’s approach
4. Lifetime spectroscopy for the determination of the defect parameters ..
Conclusion
General conclusion and perspectives
Appendix A: Correction of ECV profiles measured on ex-situ doped poly-Si contacts
Appendix B: Fitting procedure of the XPS Si 2p spectral range of a SiOx layer
List of publications
Résumé des travaux
References